USING GRIME S MATHEMATICAL MODEL TO DEFINE ADAPTATION STRATEGY OF VASCULAR PLANTS IN THE NORTH OF RUSSIA A.B. Novakovskiy, Y.A. Dubrovskiy. S.P. Maslova, I.V. Dalke Institute of Biology, Komi Science Centre, Russian Academy of Science novakovsky@ib.komisc.ru
Vegetation functional types S S-stress-tolerant (High stress, low disturbance Stress C R C-competitor (Low stress and disturbance) Disturbance R-ruderal (Low stress, high disturbance)
CSR classification (Secondary and tertiary strategy) C (2, -2, -2) C/CR C/CS C/CSR R/CR CR (0, -2, 0) R/CSR CR/CSR CSR (0, 0, 0) SR/CSR CS/CSR S/CSR CS (0, 0, -2) S/CS R (-2, -2, 2) R/SR SR (-2, 0, 0) S/SR S (-2, 2, -2)
The original model R Regression equations (separate for all C, S, R axes) C = + 2 2 2.5 + 0.118* CH 0.076* LS S = + 2 2 2 2 1.485 0.79* CH + 0.05* DMC 0.129* SLA 0.082* LS = 2.5 0.158* DMC + 0.31* FP + 0.3* FP 0.327*ln( LDW ) + 0.639* SLA CH Canopy Height (cm) LDW Leafs Dry Weight (mg), DMC Dry Matter Content (%) SLA Specific Leaf Area (Area / Dry Weight) (mm 2 /mg) LS FP FS Lateral Spread (six point classification) Flowering Period (count of months) Flowering Start (month number) Hodgson J.G., Wilson P.J., Grime J.P., Thompson K. Allocating C-S-R plant functional types: a soft approach to a hard problem // Oikos. 1999. Vol. 85. P. 282-294.
Goals and objectives Goal is to develop an approach to define functional types of vascular plants in the North of Russia based on the Grime s mathematical model Objectives: Select several herbaceous species with clear position in CSR classification and measure all variables which are used in the Grime's model Estimate relationships between these variables and functional types Validate the original mathematical model Adjust this model to our conditions and validate it
Sampling sites Places of data collecting: the Ilych River basin part of the Pechoro-Ilychsky Nature Reserve included in 1995 in the list of world heritage objects of UNESCO
Studied species Species Forest plants Melampyrum pratense Oxalis acetosella Trientalis europaea Meadow plants Chamaenerion angustifolium Ranunculus repens Rumex acetosella Marsh plants Andromeda polifolia Rubus chamaemorus Etc. Funct. type R/SR S/CS S/CSR C CR SR/CSR S/SC SC/CSR Etc. About 40 dicots herbaceous plants of different functional types from different communities were analyzed. Functional types were defined according to the Graim s classification
C-Axis (competitiveness) Competitiveness is the relatively ability of the plants to capture different resources such as light or mineral nutrients. Usually this is achieved by large physical dimensions. Canopy height (cm) Chamerion angustifolium Artemisia vulgaris
S-axis (stress-tolerance) S-type is associated with low nutrient regime, low biomass, longer life cycle with mostly vegetative reproduction Canopy height (cm) Pyrola rotundifolia Rubus saxatilis Lateral spread
R-axis (ruderality) Ruderal plants are usually characterized by short life cycle and seed reproduction Lateral spread Melampyrum pratense Taraxacum officinale Dry matter content
Distance 3 steps Distance - Original value - Model position Distance 1 step Trientalis europaea Original position S/CSR Model result CR Melampyrum pratense Original position R/SR Model position SR
Validation of the original model Species Funct. type (original) Funct. type (model) Distance Achillea millefolium CSR CR 2 Antennaria dioica SR/CSR R/CR 2 Artemisia vulgaris C/CR CR 1 Geranium sylvaticum CSR C/CR 2 Melampyrum pratense R/SR R/CR 3 Polemonium caeruleum CSR CR 2 Ranunculus repens CR CR 0 Rubus chamaemorus SC/CSR C/CR 2 Rubus saxatilis S/CSR C/CR 3 Rumex acetosella SR/CSR R/CR 2 Stellaria holostea CSR CR 2 Taraxacum officinale R/CSR CR 1 Thalictrum minus S/CSR CR 3 Trientalis europea S/CSR C/CR 3 Trollius europeus SC/CSR CR 3 Valeriana wolgensis CSR CR 2 Vicia sepium C/CSR C/CR 1 Viola palustris SR/CSR CR 2
Adjusted model C-axis S-axis R-axis B t p B t p B t p Constant -.462 -,053,958-1,86 -,182,859 3,317 2,449,021 CanopyHeight.896 1,971,077 -,948-1,768,107 -,001 -,844,406 DryMatterContent -,715 -,968,356,763,875,402 -,072-2,863,008 FloweringPeriod -,005 -,016,988,189,467,650 -,143-1,003,324 LateralSpread,552 1,699,120 -,515-1,343,209 -,298-3,027,005 LeafDryWeight -,061 -,142,890,120,236,818 -,002-1,420,167 SpecificLeafArea -,289 -,508,622,403,601,561 -,029-2,415,023 FloweringStart,126,311,762 -,014 -,029,977,077,395,696 B model coefficients, t Student s t-test value, p significance level Coefficient with p<0.1 have been marked (for S-axis the lowest p-value was marked)
Validation of the adjusted model Species Funct. type (original) Funct. type (tuned model) Distance Achillea millefolium CSR SC/CSR 1,0 Antennaria dioica SR/CSR SR/CSR 0,0 Artemisia vulgaris C/CR SC 2,0 Geranium sylvaticum CSR C/CSR 1,0 Melampyrum pratense R/SR SR 1,0 Polemonium caeruleum CSR CR/CSR 1,0 Ranunculus repens CR SR/CSR 2,0 Rubus chamaemorus SC/CSR SC/CSR 0,0 Rubus saxatilis S/CSR C/SC 2,0 Rumex acetosella SR/CSR C/CSR 2,0 Stellaria holostea CSR C/SC 2,0 Taraxacum officinale R/CSR SC/CSR 2,0 Thalictrum minus S/CSR SC/CSR 1,0 Trientalis europea S/CSR S/SC 1,0 Trollius europeus SC/CSR SC/CSR 0,0 Valeriana wolgensis CSR C/CR 2,0 Vicia sepium C/CSR C/SC 2,0 Viola palustris SR/CSR S 2,0
Conclusions and directions for further research Different types of functional types are characterized by different values of variables. The most significant relationship were shown by canopy height for C and S axis, and Dry Matter Content, Lateral Spread and Specific Leaf Area for R- axis Verification of the original model showed a low convergence, especially for S-axis. Verification of the adjusted model showed more accurate results. To improve the model convergence we will involve additional information (for example concentrations of nitrogen and carbon in species). Also we will try to use data sets from international databases (LEDA, TRY-DB).
Our team Svetlana Maslova Igor Dalke Yuriy Dubrovskiy Our studies were supported by the Russian Foundation for Basic Research (13-04-98829) and Government of the Komi Republic